Skip to content
AIGP

RFC-038: Domain of Concern Registry — AIGP Dialects as First-Class Artifacts — 10. The Scientific Parallel: Dialects as Research Protocols

AIGP Specification › RFC-038: Domain of Concern Registry — AIGP Dialects as First-Class Artifacts › 10. The Scientific Parallel: Dialects as Research Protocols

← 9. Dialect Composition and Inheritance · Section index · 11. Registry Governance →

10. The Scientific Parallel: Dialects as Research Protocols

10.1 Why This Analogy Is Structural, Not Metaphorical

A research protocol in clinical medicine is not a suggestion. It is the complete, declared, reproducible description of:

  • What is being measured (construct)
  • How it is measured (instruments, methods)
  • What evidence is admissible (inclusion/exclusion criteria)
  • Who may conduct the observation (qualified investigators)
  • What thresholds determine significance (statistical criteria)
  • What conditions invalidate the result (confounders, exclusions)
  • How the protocol is versioned when knowledge evolves (amendments)

An AIGP dialect is structurally identical:

Research Protocol Element AIGP Dialect Element
Study objective Concern class
Outcome measures Mediation Observation Model variables
Measurement instruments Extraction methods
Inclusion/exclusion criteria Evidence admissibility rules
Principal investigator qualifications Observer accreditation requirements
Statistical analysis plan Calculation semantics (RFC-036)
Significance thresholds Gap thresholds and hard stops
Protocol amendments Dialect versioning (§5)
Multi-site coordination Registry distribution and subscription
Informed consent Dialect subscription as governance contract

10.2 What This Gives Us

  1. Reproducibility — Two organizations subscribing to the same dialect version will observe using the same apparatus. Their results are comparable not because they agreed on outcomes, but because they agreed on methodology.

  2. Peer review — A published dialect can be reviewed, challenged, and improved by the community. The warrant and backing of each variable (Toulmin, RFC-035 §4.8) are public arguments subject to scrutiny.

  3. Progressive refinement — Dialect versions evolve as empirical evidence accumulates. What was a PATCH-level extraction improvement today may become a MINOR-level variable addition tomorrow as the community discovers new observable indicators.

  4. Institutional trust — An organization can declare “we govern to autonomous_systems@2.1.0” and a regulator, auditor, or partner can independently verify what that means by resolving the dialect from the registry.

  5. Falsifiability — A dialect that produces consistently uninformative verdicts (all MATCH, no signal) is empirically falsifiable as an observation apparatus. Its variables may not actually capture the construct they claim to measure. This is detectable from longitudinal verdict data and triggers dialect evolution.

← 9. Dialect Composition and Inheritance · Section index · 11. Registry Governance →